The ISME Journal https://doi.org/10.1038/s41396-018-0160-7
ARTICLE
Taxonomic diversity of fungi deposited from the atmosphere Cheolwoon Woo1 Choa An1 Siyu Xu1 Seung-Muk Yi1,2 Naomichi Yamamoto ●
●
●
●
1,2
1234567890();,:
1234567890();,:
Received: 16 January 2018 / Revised: 2 May 2018 / Accepted: 7 May 2018 © The Author(s) 2018. This article is published with open access
Abstract Fungi release spores into the global atmosphere. The emitted spores are deposited to the surface of the Earth by sedimentation (dry deposition) and precipitation (wet deposition), and therefore contribute to the global cycling of substances. However, knowledge is scarce regarding the diversities of fungi deposited from the atmosphere. Here, an automatic dry and wet deposition sampler and high-throughput sequencing plus quantitative PCR were used to observe taxonomic diversities and flux densities of atmospheric fungal deposition. Taxon-specific fungal deposition velocities and aerodynamic diameters (da) were determined using a collocated cascade impactor for volumetric, particle-size-resolved air sampling. Large multicellular spore-producing dothideomycetes (da ≥ 10.0 μm) were predominant in dry deposition, with a mean velocity of 0.80 cm s–1 for all fungal taxa combined. Higher taxonomic richness was observed in fungal assemblages in wet deposition than in dry deposition, suggesting the presence of fungal taxa that are deposited only in wet form. In wet deposition, agaricomycetes, including mushroom-forming fungi, and sordariomycetes, including plant pathogenic species, were enriched, indicating that such fungal spores serve as nuclei in clouds, and/or are discharged preferentially during precipitation. Moreover, this study confirmed that fungal assemblage memberships and structures were significantly different between dry and wet deposition (P-test, p < 0.001). Overall, these findings suggest taxon-specific involvement of fungi in precipitation, and provide important insights into potential links between environmental changes that can disturb regional microbial communities (e.g., deforestation) and changes in precipitation patterns that might be mediated by changes in microbial communities in the atmosphere.
Introduction The kingdom Fungi contains several million estimated species of molds, yeasts, mushrooms, and other life forms [1, 2]. Fungi prevail in the pedosphere [3, 4], from which spores, hyphae, fragments, and other propagules are released into the global atmosphere, with an estimated annual emission of 28–50 Tg [5, 6]. The emitted fungal particles are
These authors contributed equally: Cheolwoon Woo and Choa An Electronic supplementary material The online version of this article (https://doi.org/10.1038/s41396-018-0160-7) contains supplementary material, which is available to authorized users. * Naomichi Yamamoto
[email protected] 1
Department of Environmental Health Sciences, Graduate School of Public Health, Seoul National University, Seoul 08826, Republic of Korea
2
Institute of Health and Environment, Seoul National University, Seoul 08826, Republic of Korea
involved in climate systems by serving as ice-forming nuclei (IN) and cloud condensation nuclei (CCN) [7–9], and/or by absorbing and reflecting solar and terrestrial radiations [10, 11]. Aerosol particles, including biological particles, are important since they provide the surface area on which ice formation and cloud condensation are catalyzed in the atmosphere. The emitted particles eventually settle on the surface of the Earth by sedimentation (dry deposition) and precipitation (wet deposition). Such aeromycological processes contribute to global biogeochemical cycles, including the nutrient and hydrological cycles [12]. Atmospheric transport is also important for microbial migration and colonization [13, 14]. Thus, fungi in the atmosphere are key components of global ecological systems. Particle size is expected to affect flux densities of fungal deposition from the atmosphere. Aerodynamic diameter (da) is a commonly used size measure for airborne particles. This is defined as the diameter of a sphere that has a density of 1 g cm–3 with the same settling velocity as the velocity of a particle of interest. Aerodynamic diameters of atmospheric fungi are taxon-dependent [15, 16], and taxon-dependent dry deposition velocities of atmospheric fungi are therefore
C. Woo et al.
expected, with larger velocities for large multicellular sporeproducing dothideomycetes, for example. Particle size is important not only for dry deposition but also for wet deposition of biological particles. For instance, airborne particles with da > 2 μm serve as effective CCN that can efficiently collide and coalesce with smaller droplets, and therefore contribute to cloud precipitation [17, 18]. Indeed, literature reported that large pollen grains served as effective CCN in the atmosphere [19]. Knowledge is scarce regarding the diversities of fungi deposited from the atmosphere [20]. Petri dish trapping was traditionally used to study dry deposition. Previous studies based on analyses of settled, cultured colonies on Petri dishes containing nutrient media showed that large multicellular spore-producing dothideomycetes were abundant in dry deposition from the atmosphere in the UK [21] and Nigeria [22]. Rainwater was typically used for analysis of wet deposition [23–25]. One such study detected plant pathogenic Fusarium species in the atmosphere of the southeast coast of Spain using analyses of culturable fungi in rainwater samples [23]. However, these culture-based techniques are limited in their ability to detect fastidious species that are difficult to raise in culture for identification and offer limited taxonomic coverage. Previous research, for instance, reported that only 10–40% of atmospheric fungi were culturable [26], of which about 20% were sterile fungi or were mycelia sterilia that were not identifiable because of their slow-growing mycelia and lack of sporulation on conventional media [27]. Basidiomycetes were rarely observed in culture-based aeromycological studies [27, 28], indicating that they were not amenable to this approach. However, Hassett et al. [7] showed that basidiospores formed effective cloud nuclei, highlighting the need for accurate atmospheric detection methods that encompass non-culturable species. The goal of this study was to examine taxonomic diversity of fungi deposited from the atmosphere. Specifically, we analyzed differences in fungal assemblages between dry and wet deposition. Here, we used DNA-based methods to analyze atmospheric fungal species regardless of cultivability. These culture-independent methods can thus reveal accurate pictures of the atmospheric fungal assemblages, including basidiomycetes [15, 29], which are thought to play significant roles in global ecological systems.
Materials and methods Dry and wet deposition sampling Dry and wet deposition samples were collected using a custom-made automatic dry and wet deposition sampler [30] (Supplementary Fig. S1) on the roof (~20 m above ground) of a building in Seoul in South Korea (37°27′55.0″
N; 126°57′17.7″E), which is situated in the humid continental climate zone, from May to November 2015. The sampling point is surrounded by the hilly, forested area of the southern outskirts of the Seoul Special Metropolitan City. Ambient temperature, relative humidity, wind velocity, and precipitation during the sampling periods are presented in Supplementary Table S1. Each of the dry and wet deposition samplings was done in duplicate and continued for 1 month. The dry deposition samples were collected on 47 mm diameter quartz fiber substrates (QR-100; Advantec Tokyo Kaisha, Ltd, Tokyo, Japan) with its surface coated with 1 mL mineral oil (CAS no. 8042-47-5; Daejung Chemicals & Metals Co., Ltd, Siheung, South Korea) to minimize particle re-entrainment. Each of the wet deposition samples was collected in a polypropylene funnel (with its effective collection area of 167 cm2) connected to a 1 L polypropylene bottle by a Teflon adaptor, recovered immediately after each precipitation that typically continued for one to several days, and stored in 250 mL polypropylene bottle(s). In total, 14 dry and 14 wet deposition samples were collected (Supplementary Table S2). The samples were kept at −20 °C until DNA extraction.
Air sampling Each of the air samples was collected for a duration of 1 month and co-located on the same roof for deposition samplings. The sampling in August failed due to an intense rainfall event. An eight-stage non-viable Andersen sampler (AN-200; Sibata Scientific Technology Ltd, Tokyo, Japan) was used to collect airborne particles on 80 mm diameter glass fiber substrates (QR-100; Advantec Tokyo Kaisha, Ltd, Tokyo, Japan) with an air flow rate at 28.3 L min−1. No mineral oil was applied onto collection substrates for air sampling. The cutoff aerodynamic diameters at 50% collection efficiency were 0.43–0.65, 0.65–1.1, 1.1–2.1, 2.1–3.3, 3.3–4.7, 4.7–7.0, 7.0–11, and >11 μm. Substrates loaded to the stages of da = 0.43–0.65, 0.65–1.1, and 1.1–2.1 μm were not analyzed due to inadequate biomass [15]. A total of 30 air samples were collected (Supplementary Table S2). The samples were kept at −20 °C until DNA extraction.
DNA extraction DNA was extracted using a protocol reported elsewhere [31]. For dry deposition samples, one quarter fraction of each sampled substrate was used for DNA extraction. Each wet deposition sample was filtered to recover particulate matters, including fungal spores, onto a 47 mm diameter filter using a sterile MiroFunnelTM Filter Funnel (0.45 µm pore size, GN-6 Mericel® white gridded membrane; Pall Corporation, NY, USA) of which one quarter area fraction was used for DNA extraction. One-eighth area fraction of
Taxonomic diversity of fungi deposited from the atmosphere
each air sample substrate was used for DNA extraction. A PowerMax® Soil DNA Isolation Kit (Mobio Laboratory, CA, USA) was used with additional 0.1 mm diameter glass beads (300 mg) and 0.5 mm diameter glass beads (100 mg) for 3 min sample homogenization by a bead beater (BioSpec Products, OK, USA). After homogenization, DNA was extracted, purified, and eluted into 30 μL TE (10 mM Tris–HCl, 1 mM EDTA, pH = 8.0).
determined based on 4000 sequences subsampled from each library. The β diversity analyses were conducted based on Jaccard indices and Yue and Clayton theta similarity coefficients as measures of fungal assemblage memberships and structures, respectively. P-test analysis showed that the variability in characterizing the fungal assemblage structures within a sample duplicate was significantly smaller than the differences across the samples (ParsScore = 13, p = 0.0421) (Supplementary Fig. S2a).
DNA sequencing Quantitative PCR DNA libraries were constructed according to a protocol reported elsewhere [32]. Briefly, the fungal internal transcribed spacer 1 (ITS1) region was amplified with universal fungal primers ITS1F and ITS2 [33, 34] along with adapter sequences for Illumina MiSeq. One of the duplicates was not PCR-amplifiable in three of seven dry deposition samples, resulting in four of seven pairs of dry and wet deposition sample duplicates available for subsequent analyses. The purified amplicons were indexed using the Nextera XT Index kit (Illumina, Inc., CA, USA). Each of the purified indexed amplicons was normalized to 4 nM in 10 mM Tris–HCl (pH = 8.5) and pooled with an internal control PhiX (30%). The pooled, heat-denatured amplicons were loaded to a v3 600 cycle-kit reagent cartridge (Illumina), and 2 × 300 bp pairedend sequencing was performed by Illumina MiSeq. Raw sequence data were submitted to the NCBI Sequence Read Archive under project number SRP119277.
DNA sequence processing and analyses The full sequence processing protocol is available elsewhere [32]. Briefly, the paired-end reads were joined with a minimum allowed overlap of 10 bp in QIIME version 1.8.0 [35]. Chimeric sequences were excluded using the chimera.vsearch command [36] in mothur v.1.39.5 [37] against the uchime_reference_dataset_ITS1_28.06.2017.fasta database [38]. A total of 960,088 high-quality sequences were obtained from 55 libraries, with mean lengths ranging from 267 to 318 bp (Supplementary Table S2). Each of the resultant sequences were taxonomically assigned using BLASTN2.2.28+ against the UNITEdatabaseinFHiTINGSformat20-11-2016release. fasta database and classified using FHiTHINGS version 1.4 [39]. In addition, each sequence was BLASTN-searched against the latest version of the fasta file containing sequences of unidentified species hypotheses, i.e., top50_release_01.12.2017.fasta [40], for the possible presence of such rare species in the atmosphere. For taxonomic assignment, the sequences were not clustered into operational taxonomic units (OTUs), since no consensus threshold was available for fungal ITS [41]. For diversity analyses, the sequences were clustered into OTUs at 97% sequence similarity [4, 15]. For the α diversity analysis, the numbers of observed OTUs were
Quantitative PCR (qPCR) was performed to quantify gene copy numbers (GCN) of fungal ITS1 using a method reported elsewhere [32]. The same primers were used for sequencing (ITS1F and ITS2). Briefly, each of the 20 μL reaction mixtures included 1× Fast SYBR Green Master mix reagent (Clontech Laboratories, Inc., CA, USA), 10 μM forward and reverse primers, and 1 μL template DNA. Quantitative PCR was performed using a 7300 Real-Time PCR System (Applied Biosystems, Inc., CA, USA). Standard curves were generated against known concentrations of ITS1 amplicons by conventional PCR with the primers ITS1F/ITS2 and a template DNA extracted from Aspergillus fumigatus ATCC MYA4609. Each assay was in triplicate. No inhibition was found, in line with the previously reported method [31]. Briefly, no inhibition was confirmed in that we observed no significant difference in quantitated results of 106 GCN A. fumigatus DNA standard with and without spiking 1 μL DNA extract from each of air or deposition samples in question. Extraction efficiency of DNA was assumed to be 10% [31]. The reproducibility of measurements of fungal concentrations in biologically duplicated dry and wet deposition samples is shown in Supplementary Fig. S2b. Reproducibility was assessed in terms of cumulative coefficient of variation [42].
Data processing and statistical analyses Taxon-specific fungal concentration was determined by multiplying the sequencing-derived relative abundance of a taxon by the universal fungal qPCR-derived total fungal concentration [16, 32, 43]. Geometric mean of aerodynamic diameter (dg) of a given particle size distribution was computed by using the GM calculator version 1.0 [16]. Arithmetic mean of dg of all sampled months was used as a representative dg for each taxon. The dry deposition velocity (Vd) was calculated by the following equation: , 6 X 5 6 X X Vd ¼ Nj;i Fj ð1Þ j¼1 i¼1
j¼1
C. Woo et al.
where Nj,i is the airborne fungal concentration (GCN m–3) in the ith particle size interval measured for the jth month by the Andersen sampler, and Fj is the dry deposition flux density (GCN cm–2 month–1) measured for the jth month by the dry deposition sampler. August 2015 data were excluded as air sampling failed. The calculated dry deposition velocities were compared with the Stokes terminal gravitational settling velocities (VStk) given by the following equation: VStk ¼
ρ0 da2 g 18η
aerodynamic diameters (dg) of 7.32 and 4.34 μm, respectively. The dg values of selected genera are listed in Supplementary Table S3, and their taxonomic information is available in Supplementary Table S4. Genera were shown if they were detected in all sampled months and each month had more than four sequence reads from the samples of all particle size intervals combined. Rhodotorula was not detected in air samples collected in June, but was included due to its importance in dry and wet deposition (see the next section). The concentrations of selected genera are shown in Fig. 2. A total of 131 sequences from the air samples were matched with the sequences of four unidentified species hypotheses in the list of the top 50 most wanted fungi with the alignment length longer than 250 bp and 100% identity (Supplementary Table S5). These species hypotheses were SH027064.07FU, SH468151.07FU, SH455726.07FU, and SH459716.07FU.
ð2Þ
where ρ0 is the standard particle density (=1.0 g cm−3), g is the acceleration of gravity (=980 cm s−2), and η is the viscosity of air (=1.8 × 10−5 Pa s). The comparison of the calculated dry deposition velocity with the Stokes velocity gives insights into behaviors and fates (e.g., removal by wet deposition) of fungal particles in the atmosphere. Mothur v.1.39.5 was used to perform P-tests to compare fungal assemblage structures and memberships. SAS version 9.4 (SAS Institute Inc., NC, USA) was used for a paired t-test to compare the taxonomic richness between dry and wet deposition.
Dry and wet deposition More than 99% of the sequences from deposition samples were found to represent the Ascomycota and the Basidiomycota, with a larger relative abundance of Ascomycota in dry deposition (84%) than in wet deposition (35%). The annual mean flux densities of dry and wet deposition were 240,000 and 1,490,000 GCN cm–2 month–1, respectively, for all fungal taxa combined. Dothideomycetes, Tremellomycetes, and Microbotryomycetes were the three most abundant classes in dry deposition (Fig. 3a), with mean flux densities of 142,000, 30,000, and 27,000 GCN cm–2 month–1, respectively. Agaricomycetes, Microbotryomycetes, and Sordariomycetes were the three predominant classes in wet deposition (Fig. 3a), with mean flux densities of 804,000, 401,000, and 144,000 GCN cm–2 month–1, respectively. Figure 3b shows principal coordinate analysis plots of fungal assemblage memberships and structures based on OTUs of fungal ITS1 at 97% sequence similarity, demonstrating that
Results Air
Concentration, ΔN/Δlog da (GCN m-3 μm-1)
More than 99% of the sequences from air samples were found to represent the Ascomycota and the Basidiomycota, with the relative abundance of Ascomycota increasing with increasing aerodynamic diameter, i.e., 48, 47, 67, 87, and 89% for da = 2.1–3.3, 3.3–4.7, 4.7–7.0, 7.0–11, and >11 μm, respectively. The two most abundant classes were Dothideomycetes and Agaricomycetes (Fig. 1), with mean annual concentrations of 71,000 and 35,000 GCN m–3, respectively. Dothideomycete particles were larger than Agaricomycete particles, with geometric means of
8.0E+5
Agaricomycetes
Eurotiomycetes
Tremellomycetes
Others
Dothideomycetes
Sordariomycetes
Leotiomycetes
Dacrymycetes
June
May
July
October
September
November
6.0E+5 4.0E+5 2.0E+5 0.0E+0
2
10
20 2
10
20 2
10
20 2
10
20 2
10
20 2
10
20
Aerodynamic diameter, da (μm)
Fig. 1 Particle size distributions of atmospheric fungal concentrations, in terms of gene copy number (GCN) of fungal ITS1, measured by the Andersen sampler in Seoul in South Korea from May to November 2015. Monthly results are shown, except for August when sampling failed
Taxonomic diversity of fungi deposited from the atmosphere Concentration (GCN m–3)
a
b Irpex Schizophyllum Stereum Psathyrella Phanerochaete Aspergillus Phlebia Bjerkandera Sistotremastrum Hyphodontia Mycoacia Trechispora Coprinellus Cerinomyces Perenniporia Austroafricana Mycosphaerella Trametes Peniophora Paraconiothyrium Periconia Nigrospora Botrytis Trichomerium Cercospora Stagonosporopsis Aureobasidium Alternaria Epicoccum Phaeococcomyces Cladosporium Penicillium Phlebiopsis Daedaleopsis Chaetomium Didymella Coprinopsis Peroneutypa Hannaella Cryptococcus Peniophorella Phomatospora Antrodia Pseudovalsaria Gibberella Knufia Arthrinium Coprinus Rhodotorula Collophora Curvularia Selenophoma Trichoderma Filobasidium Nothophoma Alatosessilispora Sistotrema Wallemia Tyromyces Cabalodontia Junghuhnia Phellinus Ochrolechia Stemphylium
10 100 1000 10000 100000
2.1–3.3 3.3–4.7 4.7–7.0 7.0–11 11
